Hacker News Reader

Comment by danieltillett | Hacker News Reader

There is a third way which is put the company into hibernation. I was faced with this with my startup a bit more than 10 years ago now. I ran out of runway so I laid everyone off, paid the bills, and got a job. I then bought out everyone else, worked part time on the business and built it back up over then next few years to the point where I could return full time. I could have started a new business, but I believed that there was a lot of value in the old business [1] which proved to be correct.

1. Some caveats here. Firstly, I did not have many people to buy out and they were willing to sell at a reasonable price. Secondly, my business is in biotech/bioinformatics and we had put a lot of resources into R&D. This R&D had real value that could be used to bring the business back to life.

View on HN

Did you have any licensed IP, and if so, how did you deal with that during the hibernation period?

Nothing that I could not re-license again later. I lost the patents in progress as I didn’t have the resources to continue with them, but this was actually a blessing in disguise. Rather than protecting the IP by patents I was forced to concentrate on trade secrets which if you have the choice is a much better way to go.

If anyone is interested in learning more I wrote a rather long blog post about the whole experience [1].

1. http://www.tillett.info/2015/06/24/why-i-kept-my-startup-in-...

I was in a situation where angry clueless investors wanted to keep the software as an asset they were entitled to, but it was close to worthless without the engineering knowledge. They tried to shop the MVP around but got nowhere. If I could have kept it I would have tried to keep it running on the side.

Sure if your investors are going to be idiotic about the situation then this might not be possible, but it is an option worth exploring and in my case it worked out well.

One of the best comments I ever heard was from a Windows engineer talking about a source code leak several years back: "Sure, but how are they going to build it?"

Surely this developer meant that the leaked code had many dependencies on libraries that were proprietary to M$. That would make it difficult to build but this scenario is likely unique to partial code leaks at large companies.

I read (old new thing probably -- a long time ago) that windows had circular dependencies until relatively recently. A large-ish team spent half a decade untangling it. Good luck building 40mm lines of source for a 25 year old project without detailed instructions...

Even a complete code leak can be near unbuildable without knowledge of how to set up a build environment. A number of years back a product i work on was so difficult to build it was a two day effort to set up a new developer workstation with the help of the existing team. Someone with just a full code dump would have taken weeks.

This is what I like about non-SaaS models like desktop or server software -- they can have zero burning rate. Even if you're out of money you can stop developing new versions of the product, fire everyone including yourself, and simply keep your website alive (which costs pennies). Your customers keep using your product and you still have a chance to get new ones.

Yes this is true. For many years my total expenses were under $1000 a year.

Thank you for your thoughtful and thorough answer about your experiences.

Quick follow-up question, I followed up on your Bioinformatics company and seems like (pls correct if I am wrong) your product-line revolves base-calling and resolving genome assemblies from Sanger Sequencing.

How did you guys adapt to the constant changes in sequencing platforms: first generation of sequencers (Sanger) to NGS (Illumina, 454, PacBio)? Especially when a lot of sequencing centers retired a lot of Sanger facilities and staff?

This would require an very long answer, but the short version is we only concentrate on Sanger sequencing. It is an open secret that Sanger sequencing never went away. The large genome centers sold off all their sequencers and they were bought by small facilities and companies around the world. These are all still being used.

Sanger sequencing is far from dead as none of the NGS sequencers are as good at what Sanger sequencing excels at which is sequencing short sections of DNA. This demand is still there.

Reply on news.ycombinator.com